How to Motivate People with Leaderboards
by Omar Ganai and Steven M. Ledbetter
“It is an invariable principle of all play, finite and infinite, that whoever plays, plays freely. Whoever must play, cannot play.” — James P. Carse, Finite and Infinite Games
Leaderboards have become a common behavior change design pattern. Originally included in video games, leaderboards have started to infiltrate “gamified” products such as…
Nike+ (by Nike Better World), a fitness wearable device, which motivates users to exercise more by competing against their friends for activity points.
Opower, an energy tracking platform, which motivate users to reduce their energy use by showing how much energy they use compared to their neighbors.
GitHub, a project management platform for programmers, which motivates users to contribute to projects more by publicly displaying the number of contributions to projects by made by individual community members.
Product managers and designers use leaderboards to motivate people to use their products more. The idea is that leaderboards are motivating because people like to win and winning is motivating because it gives us status in our communities. But some people don’t want to win; they want to avoid losing.
Competition: A Problem of Competence
“Wanting to win” versus “wanting to avoid losing” is a subtle yet crucial distinction; Murayama and Elliot’s (2012) set of meta-analyses found the effects of competition depend on this distinction in the minds of players. When someone wants to perform better than others, they tend to benefit from competition. But when they want to avoid performing worse than others, competing tends to reduce their performance. Similarly, Burnette et al’s (2013) meta-analysis found that the desire to win is positively related to goal achievement, whereas the desire to avoid losing is negatively related to goal achievement (yay replication). Finally, Senko et al’s (2017) meta-analysis found that “wanting to win” improves the performance of participants only when it’s accompanied by strategies that support feelings of mastery. So “wanting to win” alone is not enough to inoculate players from the downsides of competitive social environments.
Design patterns that foster competition — like leaderboards — don’t improve motivation and performance for all users in all situations. Competition is good for motivation and achievement only when it helps users feel competent.Leaderboards are a motivation problem, and you can design for that.
Designing a Motivational Leaderboard
“A finite game is played for the purpose of winning, an infinite game for the purpose of continuing the play” — James P. Carse, Finite and Infinite Games
Self-Determination Theory posits that human beings seek and engage with activities that promise and satisfy three Basic Psychological Needs. All user interactions with an interface can support or frustrate the satisfaction of these Basic Psychological Needs (Peters, Calvo, & Ryan, 2018). So if you want to design a motivational leaderboard — and hopefully avoid accidentally demotivating a large sub-section of the people looking at it — you must design with these three needs in mind.
A great leaderboard supports competence, which is the satisfaction you get when you complete a challenging goal. You feel masterful and effective. You feel that you’re achieving things. The opposite of competence is feeling ineffective and helpless.
A great leaderboard should support relatedness, the satisfaction you get when you feel understood and liked by people you care about. You feel closer to the people you’re playing with. The opposite of relatedness is feeling rejected and disconnected.
Finally, a great leaderboard should support autonomy, the satisfaction you get when you act with a sense of personal commitment and choice. The opposite of autonomy is feeling coerced and manipulated.
Well-designed leaderboards promote the satisfaction of competence, relatedness, and autonomy for most users, most of the time. And most importantly, they don’t pit these needs against each other. For example, a risk with leaderboards is implying to players that they have to choose between competence or relatedness. A leaderboard can do this by suggesting that we are in a zero-sum competition against the people with whom we are playing. When this happens leaderboards satisfy competence for a few users at the price of frustrating relatedness for many.
Think about it like a “game night” with some friends. For most people, the point of a game night is the excuse to hang out and be with each other (relatedness). Most board games are designed with low stakes (Monopoly money not being legal tender) and cooperative play (teams, couples, etc). This is to prevent competition from pitting relatedness against competence. But if you pick a game that is designed differently, like Strip Poker or a full-on Fight Club, then you’ll most likely see more negative impacts on the motivation of your friends than positive ones. Some players would no doubt leave rather than participate and potentially wreck their friendship, and some who stayed would feel coerced into playing (both examples being the need for relatedness pitted against the need for autonomy). Even those who did play might regret it as clothes came off or noses bled.
But we can break down leaderboards into their behavior change elements and examine the impact that each element has on Basic Psychological Needs. And in doing that, show you how you can design these elements so that they satisfy competence, relatedness, and autonomy, without pitting them against each other.
The Mantra: Competence and Connection over Competition
Behavior Change Elements of a Leaderboard
A Behavior Change Technique is an “active ingredient that brings about behavior change” (Susan Michie et al. 2013). The Human Behavior Change Project has identified 93 of these “BCTs” in research, and most leaderboards are a system of four:
In turn, each of these four behavior change techniques can support or frustrate the three Basic Psychological Needs.
Let’s walk through how you can design each behavior change technique to support the Basic Psychological Needs, rather than frustrate them.
A goal is the aim of an action (Locke & Latham, 2013). Goal-setting involves giving or guiding a user toward a goal, and has become recommended as an effective building block for behavior change (Epton, Currie, Armitage, 2017).
“Get the number 1 ranking” is the default goal that is communicated by most leaderboards. No one has to say it; the design itself tells players to make “be highest on the board” their goal.
Nike+ ranks people on fitness activity points. The goal is to get the most activity points.
Opower ranks people on kWh/day of energy use. The goal is to use the least amount of energy.
Github ranks people on number of contributions to a project. The goal is to maximize your contributions.
All goals are not created equal. Some goals support Basic Psychological Needs and some goals can thwart them, depending on the person and social environment. “Run a marathon” can be a motivating goal to a middle-aged white person, whereas “run a marathon” can be a meaningless goal to a Kenyan kid struggling in poverty. Dennis Kipruto Kimetto — the 34yo world-record holder in the marathon — didn’t start running competitively until he was 27 because he didn’t think running could feed his family.
Here are some suggestions for how to design goals that support the Basic Psychological Needs of the most people instead of thwarting them, so that people will use your product long enough to actually benefit from it.
Provide a motivational “why”
Give your users a reason why putting in effort to compete on your leaderboard might be worth it, in a way that is meaningful to them (Steingut, Patall, & Trimble, 2017). From a Self-Determination Theory perspective, many products damage long-term user engagement because they provide people with The Wrong Whys.
The Wrong Whys include:
Emphasizing meaningless incentives. Example: “You should play because you’ll win all the points!”
Emphasizing boosting ego and pride. Example: “You should play to win because only the best and smartest people win!”
Emphasizing outcomes people don’t control. Example: “You should play so you remain physically attractive to others.”
All of these “whys” accidentally run the risk of making users feel stupid, isolated, and manipulated when they participate in your leaderboard.
The Right Whys include:
Emphasizing fun, enjoyment, or interest. Example: “Play because it’s fun and you’ll learn interesting facts about yourself!”
Emphasizing congruence with user personal values or desired social identity. Example: “We play because we’re a fun-loving community who help each other to become better people!”
Emphasizing autonomy. Example: “You may like playing this” instead of “You must play this”.
All of these “whys” are likely to satisfy competence, relatedness, and autonomy when users participate in your leaderboard.
Nest extrinsic goals inside of intrinsic goals
We strongly recommend giving users intrinsic goals over extrinsic goals. Intrinsic goals are about pursuing ends that are inherently valuable to us, such as having fun, achieving mastery, and building close relationships. Extrinsic goals are about pursuing instrumental outcomes such as wealth, fame, and hotness. In general, intrinsic goals…
Satisfy our Basic Psychological Needs better than extrinsic goals (Grouzet et al 2005).
Get us to put in more effort than extrinsic goals (Vansteenkiste, Simons, Lens, Sheldon, & Deci, 2004; Vansteenkiste, Simons, Soenens, & Lens, 2004).
Can feel easier to pursue than extrinsic goals (Werner, Milyavskaya, Foxen-craft, & Koestner, 2016).
Promote deeper learning compared to extrinsic goals (Vansteenkiste, Simons, Lens, Soenens, & Matos, 2005; Vansteenkiste, Timmermans, Lens, Soenens, & Van den Broeck, 2008)
Promote enhanced focus on the task at hand, compared to extrinsic goals (Vansteenkiste, Matos, Lens, & Soenens, 2007).
Perhaps most interestingly, pursuing extrinsic goals for intrinsic reasons, like aiming to make money to support your family, is more supportive of the Basic Psychological Needs than pursuing an extrinsic goal for extrinsic reasons, like aiming to make money for the sake of fame (Landry et al. 2016).
Give people intrinsic goals rather than extrinsic goals.
Example of an intrinsic goal: “Aim to improve your skills.”
Example of an extrinsic goal: “Aim to become famous.”
If you must use an extrinsic goal, nest it inside an intrinsic one.
Example: “Play to win points (an extrinsic goal because points are extrinsic to the activity), but more importantly, play to contribute to our project that is making the world a better place (an intrinsic goal because most people would agree making the world a better place is inherently valuable).”
Nest individual goals inside of team goals
Make competing more about playing with others rather than being top dog. For example, show your users how pursuing their individual goals helps other people on their team (Adam Grant, 2012). This supports relatedness because it encourages users to work with each other rather than against each other.
Nike+ comes close to pulling this off when they say, “With group goals, you can get to green together.” But then they pit competence against relatedness by saying, “Or compete individually for the top slot on the leaderboard.” They could fix this by saying something like, “Or focus on improving your personal score”, instead.
Feedback lets users know the progress they are making from Point A (where they are) to Point B (their goal). Whenever someone checks your leaderboard, they are receiving feedback.
Nike+ tells people how many activity points they’ve earned.
Opower, tells people how much energy they are using.
Github tells people the number of commits they’ve made to a project.
Well-designed feedback can promote feelings of competence and mastery. In general, the more often people check their progress, the more progress they make (Harkin et al. 2016), so it’s a good idea to design feedback that people find useful and enjoyable. Here are some tactics for doing so.
Provide feedback on meaningful standards
Provide users with feedback on standards that are meaningful to them, given their goals. Below is an example from Google’s Grasshopper. Users are being given feedback during onboarding. Specifically, it’s telling users how to make a choice (autonomy support) and letting them feel good about making the choice (competence support). It also helps users feel they are in the “right place”, which supports competence.
Provide multiple levels of feedback
Provide users with multiple ways to assess their progress. However, focus mostly on process and performance feedback because such metrics are the most under user control, and therefore have the greatest potential to support user competence.
Here are some examples of other types of feedback:
Provide juicy feedback
Juicy feedback is varied, unexpectedly excessive sensual positive feedback on small user actions and achievements (Sebastian Deterding, 2015). Juicy feedback can instill a sense of competence even without a big challenge to overcome. Here’s another example from Google’s Grasshopper:
3. Social Comparison
Social comparison is used to show users how they are doing in relation to other people. Whenever you display rankings to your users, you are employing social comparison.
Nike+ ranks people on fitness activity points, in comparison to their friends.
OPower ranks people on how much energy they use, in comparison with their neighbors.
Github ranks contributors based on how many commits they make to each project, in comparison with other project community members.
Social comparison is a tricky behavior change technique to get right. Done well, it can support competence and relatedness. Done poorly, it will make people feel ineffective and excluded. By default, people choose to compare themselves to people who are better than them on some quality (Gerber, Wheeler, & Suls, 2018). We compare upwards even when we receive negative feedback. This is surprising because you could make the argument that after we receive negative feedback, we would be more likely to compare ourselves to someone worse off than us to feel better about ourselves. But people compare upwards instead. And this upward comparison effect is stronger when we compare ourselves on ability, or on a dimension we don’t understand well. The big risk with social comparison is that people will compare themselves to people above them, resulting in reduced competence and relatedness.
Another risk with social comparison is that it “works” in the short term. Feelings of inferiority and exclusion can push people to do what you want them to do with your product. Yet this might come at the price of their long-term engagement with your product. Eventually people quit using things that make them feel chronically guilty and shameful.
Here are some suggestions to design social comparison that support Basic Psychological Needs instead of frustrating them.
Take the sting out of losing
Winning, losing, ranking, it’s almost impossible not to form an opinion on “how am I doing in comparison with others” in an competitive environment. Winning obviously feels great, but losing, not so much. Here are some ways to take the edge off losing:
Tell losers they met a competence-satisfying standard, even if they lost (Vansteenkiste & Deci, 2003).
Example: “You made more commits than 70% of the people in this project!”
Tell losers why they got the score they got (so they can learn how to improve, which supports competence and so they know the game is fair, which supports relatedness).
Example: “Looks like you were just shy on [this metric]. A little better there and you could be on top.”
Acknowledge that losing might suck (this supports autonomy because you aren’t making them feel bad about feeling bad).
Example: “Oof, so close you can taste it.”
Give losers a choice to keep playing or not (this supports autonomy; Chatzisarantis, Kee, Thaung, & Hagger, 2012). Example:
Keep the focus on “playing and connecting” rather than on winning
Keep winners and losers focused on playing well rather than achieving victory (Standage, Dude, & Pensgaard 2005) by providing some suggestions on how they can have more fun, improve, or connect with other players. Losing with a teammate is less demotivating than losing alone (Standage, Dude, & Pensgaard 2005), so emphasize connections and relationships.
Example: “Would you like to talk to other people who have been working on this problem?”
Compare people on a metric they can improve and have improved on in the past
Rank people on how often they choose to play or use your app, rather than on an outcome measure they have less control over (such as pounds of bodyweight lost). The idea is to satisfy their competence. A focus on learning and mastery limits any frustration of competence when you receive a negative social comparison (Chatzisarantis et al 2016; Kamarova et al 2017).
Example: “You’re currently on a 10-day streak. Would you like to see how other people have done at this when they were on a 10-day streak?”
Make it harder to compare
The negative effects of social comparison are less potent when people have to compare themselves to more than two options (Gerber, Wheeler, & Suls, 2018). Take all your instincts about usability and flip them on their head. For example, we know it’s hard for people to hold more than seven pieces of information. So compare them to more than 7 people. Alternatively, maybe you can use unique metrics for each person, so that the leaderboard acts more like a place for you to see the personal progress everyone is making, rather than a place to compare yourself to others.
Example: “Would you like to explore how other people are learning and progressing by looking at some of their metrics?”
4. Social Rewards
Social rewards are non-material/financial incentives given in response to a user putting in some effort. Perhaps the most successful social reward in history is the Facebook Like. Many products justify the use of leaderboards by claiming that winning will win users accolades from people they care about, which will motivate them to keep using the product.
At this point it’s well-known that rewards can undermine intrinsic motivation (Cerasoli, Nicklin, & Ford, 2014; Deci, Koestner, & Ryan, 1999), so we should use them carefully. If the behavior you want your users to do is repetitive and boring, linking rewards to effort is justified. If the behavior you want your users to do requires creativity, make rewards surprising and intrinsic.
We can design rewards to limit the damage to Basic Psychological Needs.
Make them surprising
Predictable rewards tied to effort undermine intrinsic motivation. Make rewards surprising.
Example: “You just had a 47 day streak! Here’s a prize!”
Tie them to big picture life goals
Example: “The more points you earn on our leaderboard, the more we donate to a charity of your choice!”
Use intrinsic or symbolic rewards
Example: Starbucks Rewards Card rewards people with more coffee instead of paying them cash.
It’s easy to think, “how do we motivate people?” and then think, “Leaderboards!” They’re so ubiquitous that we often don’t think of how they’re implemented and what assumed values they are silently communicating to us. We hope that by breaking them down into component Behavior Change Techniques, we’ve shown you how the parts can come together to either motivate or demotivate depending on how those parts work together to support (or thwart) Basic Psychological Needs. If you take the time to look at your own Leaderboard and think through our Mantra: “Competence and Connection over Competition”, we think you’re more likely to motivate the people using your products and less likely to accidentally make them feel like losers.
Links & Notes
View references list.